Jan Sebek1, Nathan Albin2, Radoslav Bortel3, Bala Natarajan4, Punit Prakash4. 1. Department of Circuit Theory, Czech Technical University in Prague, Praha 6 16627, Czech Republic and Department of Electrical and Computer Engineering, Kansas State University, Manhattan, Kansas 66506. 2. Department of Mathematics, Kansas State University, Manhattan, Kansas 66506. 3. Department of Circuit Theory, Czech Technical University in Prague, Praha 6 16627, Czech Republic. 4. Department of Electrical and Computer Engineering, Kansas State University, Manhattan, Kansas 66506.
Abstract
PURPOSE: Computational models of microwave ablation (MWA) are widely used during the design optimization of novel devices and are under consideration for patient-specific treatment planning. The objective of this study was to assess the sensitivity of computational models of MWA to tissue biophysical properties. METHODS: The Morris method was employed to assess the global sensitivity of the coupled electromagnetic-thermal model, which was implemented with the finite element method (FEM). The FEM model incorporated temperature dependencies of tissue physical properties. The variability of the model was studied using six different outputs to characterize the size and shape of the ablation zone, as well as impedance matching of the ablation antenna. Furthermore, the sensitivity results were statistically analyzed and absolute influence of each input parameter was quantified. A framework for systematically incorporating model uncertainties for treatment planning was suggested. RESULTS: A total of 1221 simulations, incorporating 111 randomly sampled starting points, were performed. Tissue dielectric parameters, specifically relative permittivity, effective conductivity, and the threshold temperature at which they transitioned to lower values (i.e., signifying desiccation), were identified as the most influential parameters for the shape of the ablation zone and antenna impedance matching. Of the thermal parameters considered in this study, the nominal blood perfusion rate and the temperature interval across which the tissue changes phase were identified as the most influential. The latent heat of tissue water vaporization and the volumetric heat capacity of the vaporized tissue were recognized as the least influential parameters. Based on the evaluation of absolute changes, the most important parameter (perfusion) had approximately 40.23 times greater influence on ablation area than the least important parameter (volumetric heat capacity of vaporized tissue). Another significant input parameter (permittivity) had 22.26 times higher influence on the deviation of ablation edge shape from a sphere than one of the less important parameters (latent heat of liver tissue vaporization). CONCLUSIONS: Dielectric parameters, blood perfusion rate, and the temperature interval across which the tissue changes phase were found to have the most significant impact on MWA model outputs. The latent heat of tissue water vaporization and the volumetric heat capacity of the vaporized tissue were recognized as the least influential parameters. Uncertainties in model outputs identified in this study can be incorporated to provide probabilistic maps of expected ablation outcome for patient-specific treatment planning.
PURPOSE: Computational models of microwave ablation (MWA) are widely used during the design optimization of novel devices and are under consideration for patient-specific treatment planning. The objective of this study was to assess the sensitivity of computational models of MWA to tissue biophysical properties. METHODS: The Morris method was employed to assess the global sensitivity of the coupled electromagnetic-thermal model, which was implemented with the finite element method (FEM). The FEM model incorporated temperature dependencies of tissue physical properties. The variability of the model was studied using six different outputs to characterize the size and shape of the ablation zone, as well as impedance matching of the ablation antenna. Furthermore, the sensitivity results were statistically analyzed and absolute influence of each input parameter was quantified. A framework for systematically incorporating model uncertainties for treatment planning was suggested. RESULTS: A total of 1221 simulations, incorporating 111 randomly sampled starting points, were performed. Tissue dielectric parameters, specifically relative permittivity, effective conductivity, and the threshold temperature at which they transitioned to lower values (i.e., signifying desiccation), were identified as the most influential parameters for the shape of the ablation zone and antenna impedance matching. Of the thermal parameters considered in this study, the nominal blood perfusion rate and the temperature interval across which the tissue changes phase were identified as the most influential. The latent heat of tissue water vaporization and the volumetric heat capacity of the vaporized tissue were recognized as the least influential parameters. Based on the evaluation of absolute changes, the most important parameter (perfusion) had approximately 40.23 times greater influence on ablation area than the least important parameter (volumetric heat capacity of vaporized tissue). Another significant input parameter (permittivity) had 22.26 times higher influence on the deviation of ablation edge shape from a sphere than one of the less important parameters (latent heat of liver tissue vaporization). CONCLUSIONS: Dielectric parameters, blood perfusion rate, and the temperature interval across which the tissue changes phase were found to have the most significant impact on MWA model outputs. The latent heat of tissue water vaporization and the volumetric heat capacity of the vaporized tissue were recognized as the least influential parameters. Uncertainties in model outputs identified in this study can be incorporated to provide probabilistic maps of expected ablation outcome for patient-specific treatment planning.
Authors: H Petra Kok; Erik N K Cressman; Wim Ceelen; Christopher L Brace; Robert Ivkov; Holger Grüll; Gail Ter Haar; Peter Wust; Johannes Crezee Journal: Int J Hyperthermia Date: 2020 Impact factor: 3.914
Authors: Jarrod A Collins; Jon S Heiselman; Logan W Clements; Jared A Weis; Daniel B Brown; Michael I Miga Journal: IEEE Trans Biomed Eng Date: 2019-09-05 Impact factor: 4.538